InformationWeek 500: Orion Builds Efficiency Into Eli Lilly's R&D System
Collaboration, automated data collection chores expedite delivery of new products.
Eli Lilly's scientists are working on a few thousand potential new drugs at any given time, and the company always needs to be thinking about its next potential blockbuster. Medications that make it through research and development, get regulatory approval, and find their way to consumers have patent protection only for a limited time before makers of generic drugs start selling their own versions, so Lilly must constantly innovate.
And, of course, competitors are always striving to come up with something better.
The prescription: a centralized portfolio management system that helps Lilly's managers understand the potentials and risks of all the drugs in its pipeline. The system, named Orion, tracks all of Lilly's potential drugs, letting some 350 R&D business managers analyze hundreds of factors, such as the amount of money that's been spent on testing, the length of tests, business risks, market need, and the expected return on investment if a drug makes it to market.
Managers also can view the big picture—factors such as how many studies are in progress across the portfolio and the stage of studies. These types of views let managers weight the portfolio much like a stock portfolio, so the best bets, based on probabilities, get the time and resources they need.
Lilly's IT department deployed the new system last year, replacing one that required IT employees and businesspeople to pull data on drugs from various sources, including the company's SAP ERP system, Oracle databases, and spreadsheets, and create a rudimentary portfolio view on Excel spreadsheets and Word documents.
In late 2006, the R&D department came to IT for help devising a better way. Orion, which was developed in less than a year at a cost of about $10 million, encompasses a data warehouse; Microsoft's PerformancePoint business intelligence software for delivering reports, charts, and graphics; Microsoft SharePoint as a system portal; and features and functions built by Lilly's own developers.
Orion automates the process of pulling data from the SAP ERP system and other data sources, eliminating many hours of work. One newly automated visual presentation—a "relationship diagram," which lets managers collapse charts for multiple drug projects into a program-level view—saves Lilly $500,000 a year in labor costs, the company says. Lilly now can assemble portfolio information in a matter of hours, a process that used to take a month or two, lessening the risk that data could be outdated by the time it was analyzed.
Authorized managers can tap into Orion from their PCs and view consistent information, so they can make faster, collaborative decisions on what drugs to push to market and what projects to put on hiatus. "Our ability to access, apply, and manage information through the portfolio, from basic innovation to marketplace dynamics, has to advance dramatically, and this is one of those tools that will help us do that," says CIO Mike Heim.
Deloitte consultants worked with Lilly for about two months to define the problem it wanted to solve, what types of technologies might solve it, and the business processes that had to change, says Bill Cox, manager of global medical and regulatory IT.
Lilly adopted Microsoft SharePoint as a company portal and PerformancePoint for other BI needs in late 2006, and, after comparing alternatives, decided to make Microsoft's BI product line, including SharePoint and PerformancePoint, the cornerstones of Orion. But there was plenty of custom development to do, including getting data out of the SAP ERP and Oracle databases and into a new data warehouse built on SQL Server. Lilly's IT group also had to design and build the SQL cubes from which data could be analyzed.
Lilly tapped Microsoft's consulting services for help building the data warehouse, integration, and reporting services. Two consulting firms specializing in business intelligence, Glenture and Oculus, were brought in to help build BI cubes and to turn complex data into visual graphics.
InformationWeek Tech Digest, Nov. 10, 2014Just 30% of respondents to our new survey say their companies are very or extremely effective at identifying critical data and analyzing it to make decisions, down from 42% in 2013. What gives?